Driver bias may cause significant variations in the fuel economy of a vehicle. For example, studies conducted by the U.S. Environmental Protection Agency (EPA) have shown that, even when holding the vehicle and the route constant, driver behavior may account for up to 35% variation in fuel economy (FE).
Driver assistance systems for optimizing fuel economy may be categorized into two general groups based on their mechanisms of influence. Passive assistance systems include those where only advisory feedback is provided to the driver. In active assistance systems, system automation takes over some portion of vehicle control. Some active automation technologies work relatively independent of the driver, such as adaptive cruise control, and some work in conjunction with the driver, such as an automated/automatic transmission.
Passive driver assistance schemes for fuel economy have been around for some time. Many of the original equipment manufacturers (OEMs) offer some type of driver interface for fuel economy in at least some of their vehicle models. The commercial vehicle market, on the other hand, is typically dominated by aftermarket devices, which are often coupled with telematic solutions. In recent years, nomadic devices have also emerged, boosted by smart phone technology, but their effectiveness may be hindered due to limited access to the vehicle's data bus and limited capability as a driver interface.
Existing fuel-economy interface devices may provide visual feedback of instantaneous or average fuel economy. Some may also provide a fuel efficiency “score” to the driver. The existing interface devices, however, are not noted for providing actionable information to the driver in terms of which behavior is related to poor fuel economy or how to operate the vehicle in a more fuel-efficient manner. In addition, existing fuel economy driver feedback schemes are typically based on vehicle operations rather than the driver's interaction with the vehicle operating environment. For example, an aggressive driver braking hard while tailgating may be rated the same as a conservative driver being forced to brake hard because another vehicle suddenly cut in front of him, as both will lead to the same hard braking event and the same fuel economy assessment. As a result, the fuel economy score or rating given to the driver is often confusing and misleading, which makes it difficult for both the driver to make use of the data.
Currently there is a general lack of actionable information and a poor differentiation between driver-caused and environment-caused fuel economy inefficiency. Both of these factors limit the effectiveness of existing fuel-economy driver interface technologies. In addition, since the driver's response to any passive feedback is typically going to be slow and coarse by nature, passive feedback may become distracting and ineffective in situations that require fast, frequent, or high-accuracy response from the driver.
As for active feedback technologies, examples include synthetic brake pulse or acceleration resistance, road speed governor, automated/automatic transmission, and standard or adaptive cruise control, to name a few. Some of these technologies were not intentionally introduced for the purpose of improving fuel efficiency, but have advanced in recent years to assist fuel economy determinations. For instance, current automated/automatic transmissions may have the capability of estimating road grade and load and adjusting gearshift schedule for better fuel economy.
Existing active feedback devices, however, have limitations similar to passive devices. Many existing active feedback solutions do not have the intelligence to differentiate between driver-caused and environment-caused fuel economy inefficiency. In addition, many of the control strategies employed in active assistance solutions are based only on instantaneous information, such as the load-based gearshift scheduling of many newer automated/automatic transmissions.
Achieving optimum driver behavior and powertrain operation to maximize fuel efficiency depends on the ability to estimate future conditions of the driver, vehicle and environment. Lack of knowledge of or inappropriate response to such future conditions may limit the effectiveness of vehicle automation systems. In recent years, progress in intelligent transportation systems (ITS) has increased availability of road and traffic information at both vehicle level and fleet level. Since fuel consumption is a cumulative measure over time, the optimality of instantaneous driver behavior is significantly dependent on future conditions. The availability of predictive road and traffic information from modern ITS technologies, such as global positioning systems (GPS), digital maps and radars, make it feasible and affordable to reduce driver bias through enhanced driver feedback and/or powertrain automation. Most of the early solutions have been focused on utilizing predictive topographical information to reduce fuel consumption in a cruise control context, where the vehicle is switching between full autonomy and full driver control per driver's choice. Since the fuel-efficient control strategy will only be active in the full autonomy mode, accurate, predictive sensor information is generally required to enable the system to make appropriate decisions consistently to avoid frequent driver intervention. With the help of technologies such as GPS and digital maps, reasonably accuracy information is readily available for simple and largely static environments, such as long distance travel on an open freeway. For the more complex urban environment, where future traffic conditions may be fluid and unpredictable, sensor and decision errors are inevitable and the driver is more likely to simply take over control of the vehicle, which makes cruise-control based solutions less effective.